If I had had an account averaging $1,000,000 at risk on any given day, then I could have (and did) pull out $290k per year on average.
But the amount of money at risk on a given day would probably have varied quite a bit, maybe from $500k to $1.5m.
The key word is “average”.
In short, it really worked, but it only worked because I had other investments (including a cash reserve and a mortgage)
which I could move money to and from when I added to or lightened up on my Berkshire position based on its attractiveness.
How many times have I said I learn something from Jim? Well today I learned I didn’t really know what IRR was.
My takeaways at this point, and please correct me where I am getting this wrong, are as follows.
IRR depends on the actual cash flows in and out of the account! So IRR is a metric of how good an investment is, but the “investment” it measures includes the effect of actual withdrawals and deposits. So if withdrawals and deposits are being made strategically, IRR gives a metric which is a combination of the investment itself (e.g., BRK stock and options) and the strategy for investing and disinvesting over the life of the investment. It was be as if you were in a mutual or hedge fund which would dictate when it would send you cash from your investment, and when it would demand cash from you to put in your investment, then IRR would be the proper way to calculate how well that Hedge fund was treating you. Except it wouldn’t allow you to evalue the “cost” to you of having to take cash when they wanted and having to produce cash when they wanted.
The 29% number I have been calling IRR for Saul for 1988 thru 2021 inclusive, it turns out, is an IRR. But it is the IRR where the SAUL stocks are invested in in 1988, all the money is left 100% in stock but rebalanced as Saul determines, and then removed at the end of end of 2021. So there is a cash flow in of $1 in 1988 and a cash flow out of $4,486.88 in 2021, and that is an IRR of 29%.
Since it apparently matters, I tried two other cash flow strategies, but with the same annual returns reported for Saul, to see how they might change things.
First, I figured I’ll just take the earnings (or refund the losses) every year. So my cash flows are,
$1 to get into the stocks in 1988,
an average of $0.36 a year earnings but lumpy, ranging from $2.33 one year to -$0.63 paid back in a different year.
Taking out my $1 principal plus my final earned interest in 2021.
In this case, my IRR is 27%, not that much different from 29%.
Second, I tried just taking a steady 25% of my previous years balance each year. This pulls my IRR down to 26% and also gives me a decade of about $0.06 a year to live on in the 2010-2020 range.
Third, I tried a few other interest rates and they tend to leave IRR in the 25+% range.
So for the SAUL annual returns, 1989-2021, it seems hard-ish to get IRR much lower than 25%, but of course I didn’t try any extreme tricks. So even though 25% IRR getting paid 20% of the balance per year to live on is not quite 29%, it is sustainable on the minimal testing I’ve done. No doubt you would either want to take a lot less or do some serious testing of hypothetical future returns before being that aggressive.
By the way, I am using Excel’s function to calculate IRR, not the average capital-years at risk dividing the total profits. I assume your intuitively interesting formulation and Excels are in agreement?
So regarding your own performance with BRK and long BRK calls and covered short BRK calls…
Presumably your “in” and your “out” are sufficiently different in magnitude and duration that you would need something like a peak of $10million in your strategy to average something like $1million in your strategy, which is to say in order to get $290,000 per year you would have to be deciding, most of the time, what to do with the other $9,000,000?
When you write covered calls, you presumably are booking the owned underlying shares as capital invested. So if you are really swinging a lot, you are spending a significant amount of time where you are both low on BRK stock and not writing covered calls. How do you determine when NOT to harvest time value from covered calls even though you don’t think the stock is going up? I presume a period of stock+CoveredCalls is not quite as short as being out of both, and that is how you decide?
So some low Price/Book has you long calls (and stock?),
some higher but still low P/B has you long stock but not calls,
some middling P/B has you in stock but with covered calls
some higher P/B has you out of stock and calls holding cash waiting for P/B to drop to get back in?
Anyway, I am very happy to understand IRR significantly better than I did. Thanks for that and for the insights on the moving pieces of stock and covered calls.
My own trading is in tax deferred us retirement accounts, so I don’t worry about tax per se, but I am limited to long calls long puts and covered calls and unable to write naked calls or naked puts.